Dario Amodei says Anthropic plans to fight the DOD's risk designation in court, claims the DOD's letter has a “narrow scope”, and apologizes for his leaked memo
Yesterday (March 4) Anthropic received a letter from the Department of War confirming that we have been designated …
Google and Amazon join Microsoft in saying they will keep working with Anthropic on non-defense projects after DOD designated Anthropic a supply chain risk
Google said it will continue offering Anthropic's artificial intelligence technology for clients, excluding for defense work …
Admin official: Pentagon may sever Anthropic relationship over AI safeguards; Anthropic says only mass surveillance and fully autonomous weapons are off limits
Admin official: Pentagon may sever Anthropic relationship over AI safeguards; Anthropic says only mass surveillance and fully autonomous weapons are off limits
The Pentagon is considering severing its relationship with Anthropic over the AI firm's insistence on maintaining some limitations …
This year's projected capex ramp-up will all but wipe out free cash flow for Amazon, Google, and Meta, potentially forcing stock buyback cuts or more borrowing
Big tech's dramatic ramp-up in projected capital expenditures this year will all but wipe out free cash flow for Amazon, Google and Meta Platforms.
Apple, whose capital expenditure is just a fraction of its peers, is the only Big Tech company that saw capex decline YoY in Q4, decreasing 19% to $2.37B
One of these companies is not like the others. — If you've been following the Big Tech companies' earnings reports …
Apple, whose capital expenditure is just a fraction of its peers, is the only Big Tech company that saw capex decline YoY in Q4, decreasing 19% to $2.37B
One of these companies is not like the others. — If you've been following the Big Tech companies' earnings reports …
Anthropic says Claude Opus 4.5 is “harder to trick with prompt injection than any other frontier model in the industry” but isn't “immune” to such attacks
Hayden Field / The Verge :
Anthropic launches Claude Opus 4.5, saying it is “the best model in the world for coding, agents, and computer use” and “meaningfully better at everyday tasks”
Our newest model, Claude Opus 4.5, is available today. It's intelligent, efficient …
Trump's One Big Beautiful Bill Act requires all new US border surveillance towers to be “autonomous”, effectively granting Anduril a monopoly on tower contracts
Anduril Industries is a major beneficiary of the One Big Beautiful Bill Act, which includes a section that essentially grants … Bluesky: @jesspish , @mmasnick , @eff.org , and @sam...
Microsoft says internal and external reviews found no evidence the Israeli military used its tech to harm civilians in Gaza, after months of employee protests
blogs.microsoft.com/on-the-issue... Sam Biddle / @sambiddle.com : Microsoft Investigates Microsoft, Clears Microsoft of Wrongdoing blogs.microsoft.com/on-the-issue... @jaybut707 : ...
Sources: the CFPB terminated more than 1,400 workers, leaving about 200 employees and effectively gutting the agency that Elon Musk previously said to “delete”
The Consumer Financial Protection Bureau terminated the positions of 1,400 of the agency's 1,700 employees …
An arbitrator instructs a former Meta employee to stop promoting and publishing her book alleging company misconduct; publisher Flatiron Books earlier objected
www.nytimes.com/2025/03/12/t... Josh Sternberg / @joshsternberg.com : Guess what book I just bought! — www.nytimes.com/2025/03/12/t... Karl Bode / @karlbode.com : this must be th...
An arbitrator instructs a former Meta employee to stop promoting and publishing her book alleging company misconduct; publisher Flatiron Books earlier objected
An arbitrator has instructed the book's author and its publishers to stop publishing the book, though it's unclear how much authority the arbitrator has to do so.
OpenAI says it found evidence of an AI-powered surveillance tool used by a Chinese security operation to identify anti-China posts on social media in the West
The company said a Chinese operation had built the tool to identify anti-Chinese posts on social media services in Western countries.
Google drops language from its AI Principles that said it would not pursue AI applications “likely to cause overall harm”, such as for weapons and surveillance
In 2018 the company updated its policies to explicitly exclude applying AI to weapons. Now that promise is gone.
Donald Trump says Meta has “come a long way” after the company announced it will end its fact-checking program, and that Meta's “presentation was excellent”
www.techpolicy.press/transcript- m... @futbolsound : The swagger of a dictator in full view. The media will play into it, he'll continue to be insufferable, and the billionaires w...
Ten current and former Meta employees share frustrations over Meta's surrender to the right on speech issues; one called the changes “a precursor for genocide”
“I really think this a precursor for genocide,” a former employee tells Platformer — I. The past
Meta says it's testing an “Imagine Yourself” feature announced in September that automatically generates AI images of a user and puts them into that user's feed
But to be fair, this guy did upload his image into their AI filtering doodad. So, y'know, those of us who eschew AI will not get randomly fucked with. [embedded post] Sam Biddle ...
Sources: US officials are investigating Chinese router maker TP-Link, which has an estimated 65% US market share, and could ban the sale of its routers in 2025
TP-Link is the bestselling router on Amazon—and has been linked to Chinese cyberattacks — U.S. authorities are investigating whether …